(photo by David Goehring)
A quick survey of project management software applications shows a distinct lack of AI tools for project managers. In fact, the first adoption of AI by project managers will be more generic business tools such as managing resources, automatic report preparation or assessing the external environment. These are all fairly benign when what we really want are tools that help us deliver the project within the constraints of scope, schedule, and budget. AI tools for project management need to treat the project in a holistic manner to ensure that the complexity of all the moving parts are considered together and do not create unexpected consequences.
AI project management tools require two documents and most organizations do not currently store or maintain these in an accessible way. It is also probable that many organizations do not have these documents at all.
Lessons learned. Project managers are well aware of creating a lessons learned document for projects that have gone wrong but there also needs to be a lessons learned document for projects that go well. This is the basis for supervised learning in machine learning. The characteristics of successful and unsuccessful projects can be analyzed and then used to classify new projects in order to make changes before the project begins.
Project issues log. An issue log with the issues, actions, and outcomes provides an ideal database for reinforcement learning for AI. In addition, this document is likely unique to your organization. In other words, the way the organization approaches solutions and achieves outcomes can be considered a ‘cultural’ flavour that is considered in the AI response for a specific organization. With this document an AI tool can learn how to optimize solutions to problems as they arise in an ongoing project and those solutions will be aligned with the organization’s environment, strategy and culture.
Both documents need to be captured in a database and made easily accessible for an AI tool to use in the training process. Failure to create and store these documents will delay AI implementation in your organization as well as making it more complex and expensive to implement.